Nicholas Dunzelman

Role portfolio · checked 2026-05-19

One proof route now maps to the whole visible Codex role cluster.

Turn the 28 visible Codex openings into a reviewer map: keep AI Deployment Engineer - Codex as the spearpoint, show where the same CIW receipts could travel if OpenAI redirects the conversation, and expose the lanes that are not earned yet.

28Codex jobs visible
1Primary route
6Tracked role lanes

Rule

Use without sending

The current proof is a deployer packet, not a generic job board scrape. It is strongest where OpenAI wants Codex adoption, demos, operational readiness, workshop material, security judgment, and feedback loops. Product, applied-AI, systems, cloud, and marketing roles become navigation branches only after a verified human signal changes the route.

No new application, no outbound, and no role-switch message before 2026-05-26 unless a verified reply or explicit fresh override changes the route. Use this packet only to prepare reviewer conversation paths, not to claim OpenAI interest.

Roles

Fit router

primary · fit 5/5

AI Deployment Engineer - Codex

This is the earned lane: CIW already shows field kits, demos, scorecards, security review, public receipts, and operator discipline.

Proof route
Route index -> digest -> receipt reference -> field-trial scorecard -> security packet -> objective audit.
Gap
Conventional 5+ year post-sales/customer deployment tenure remains the honest weakness, so proof must show customer-like deployment mechanics without claiming real customers.
Next safe move
Keep this as the primary post-gate route and refresh outcome evidence before any message.

strong-adjacent · fit 4/5

Deployed Product Manager, Codex

Adjacent but plausible: CIW has product judgment, field-signal simulation, route-friction analysis, collaboration lanes, and proof packaging.

Proof route
Scale engine -> feedback export -> source index -> collaboration path -> company map.
Gap
Real strategic-account ownership is not proven; treat this as a conversation lane, not the main application lane.
Next safe move
Use only if an OpenAI human asks where the proof fits beyond deployment engineering.

stretch · fit 3/5

Applied AI Engineer, Codex Core Agent

Stretch lane: the proof web shows agent-operation taste and evaluator design, but not enough model-improvement or coding-agent internals.

Proof route
Regression suite -> deployment drill -> Cookbook-style example -> scale engine.
Gap
Needs stronger direct eval fixtures, ablation logs, and coding-agent benchmark results before it can outrank the deployment route.
Next safe move
Backlog an eval-and-ablation proof node only if the current deployment route stops moving.

stretch · fit 3/5

AI Systems Engineer, Codex Agents

Technical stretch: the autonomy kernel, ledger, graph, and verifier stack prove operator-system thinking, not low-level agent runtime engineering.

Proof route
Autonomy status -> ledger -> graph -> company map -> scale engine.
Gap
Needs stronger sandbox/runtime, telemetry, and reliability artifacts before treating this as a direct route.
Next safe move
Do not apply or pitch this lane now; use it to decide what engineering proof to build next if needed.

watch · fit 2/5

Software Engineer, Codex Cloud

Watch lane: CIW shows secure proof publication and verifier discipline, not cloud-scale backend ownership.

Proof route
Security packet -> heartbeat -> public URL checks -> launch matrix.
Gap
Needs production infrastructure evidence, service reliability work, and cloud execution depth.
Next safe move
Keep as a source of proof requirements, not a current outreach lane.

not-now · fit 1/5

Product Marketing Manager, Codex

Not-now lane: CIW has technical narrative assets, but this route likely wants senior launch and GTM ownership.

Proof route
Digest -> route index -> scale engine as narrative proof only.
Gap
Insufficient senior marketing, launch ownership, and GTM management evidence.
Next safe move
Use the role as a writing-quality benchmark only; do not treat it as a target.

Sources

Current source basis

careers-search-codex

OpenAI careers search for Codex

The search result is the terrain scan: 28 Codex openings on 2026-05-19 across deployment, product, systems, cloud, applied AI, infrastructure, research, design, and marketing.

source

target-ai-deployment-engineer

AI Deployment Engineer - Codex, Remote - US

This is the spearpoint role: Codex adoption, demos, operational readiness, feedback loops, and customer-facing deployment judgment.

source

ai-systems-engineer-codex-agents

AI Systems Engineer, Codex Agents

The systems branch asks for agent harnesses, evaluation, telemetry, sandboxing, speed, reliability, and shipped coding-agent improvements.

source

applied-ai-engineer-codex-core-agent

Applied AI Engineer, Codex Core Agent

The applied-AI branch asks for hard coding tasks, evals, product intuition, tool-use changes, and tight model/product feedback loops.

source

software-engineer-codex-cloud

Software Engineer, Codex Cloud

The cloud branch points toward secure remote execution, orchestration, infrastructure reliability, enterprise controls, and production-scale developer experience.

source

deployed-product-manager-codex

Deployed Product Manager, Codex

The product branch values strategic accounts, integrations, MCP-style extension surfaces, customer loops, and field signal translated into roadmap decisions.

source

product-marketing-manager-codex

Product Marketing Manager, Codex

The marketing branch is a narrative benchmark only; it wants senior GTM ownership, not Nic's strongest technical-operator evidence.

source

openai-interview-guide

OpenAI interview guide

The guide keeps the packet grounded in projects, technical communication, interview mechanics, and evidence instead of speculative role switching.

source

Stops

Hard gates